I don't want to start another flame war.
Byu your commnents in this gorup for at
least the last 4 years you have shown a
good deal of knowledge about RF.

Your position on transformer vs resitive
hybrids confusses me. Based on what I
have read, transformer hybrids have 3 dB
less loss then resitive ones. From your
firm stance it is clear that you disagree.
Since I respect your opinion I am interested
on why you disagree.

Trying to move this to Email to avoid
wasting space and because I doubt if 1
SWL in 1000 care about this.

email
r2000swl
AT (replace with shift 2)
hotmail
dot
com

Sorry for the mangled address but Google even
hides email addresses in the text of messages.

Terry

In article .com,
wrote:

I don't want to start another flame war. Byu your commnents in this
gorup for at least the last 4 years you have shown a good deal of
knowledge about RF.

Your position on transformer vs resitive hybrids confusses me. Based
on what I have read, transformer hybrids have 3 dB less loss then
resitive ones. From your firm stance it is clear that you disagree.
Since I respect your opinion I am interested on why you disagree.

Trying to move this to Email to avoid wasting space and because I
doubt if 1 SWL in 1000 care about this.

email r2000swl AT (replace with shift 2) hotmail dot com

Sorry for the mangled address but Google even hides email addresses
in the text of messages.

We can have differing opinions on politics and religion but this is
technical factual information and it's not a waste of space if it is on
topic.

The resistors in the divider don't connect to ground or the circuit
return so they don't consume power. They connect between the signal
paths and are an impedance matching network so that the input looks like
50 ohms to the signal source with two 50 ohm loads on the output ports
of the splitter, which are in parallel. Without the impedance matching
the power splitter would look like 25 ohms and a large reflection would
take place.

The insertion loss is usually in fractions of a dB due to the fact that
they are not perfect resistors and that the path through the divider is
not a perfectly continuous 50 ohm impedance path. Any deviation in a
signal paths impedance causes a power reflection back toward the signal
source and so is a power loss. This extra loss is usually less than 1 dB.

An insertion loss figure of merit is usually 1.5dB for anything in the
signal path.

Some manufactures specify their parts in division of power (3dB) and
some in voltage (6dB) in the case of a 1 to 2 splitter. In a 1 to 2
splitter that means that the power will be 3dBm less at either output
port than the input port and that the voltage swing with be 6dBV less at
either output port than the input port.

The insertion loss will be the extra loss over the power division loss
(3dBm or 6dBV) that is understood as the power division function.

I looked around on the net and did not come up much but here are some
links to applications notes on the Mini-circuits web site. They use
transformers in their splitters. You can compare their insertion loss
figures with the much wider bandwidth microwave resistive splitters on
other companies web sites.

Power splitter and combiners:

Understanding power splitters
http://www.minicircuits.com/appnote/psc2-2.pdf

Frequently asked questions
http://www.minicircuits.com/appnote/pwr2-4.pdf

--
Telamon
Ventura, California

I will study both pdfs and see if I can come up with a couple of
experiments
to see if I understand this.

I am picking up a Pasternak PE2066 DC~6GHz with "N" connectors
from a friend this afternoon.

I have a 1W 7MHz/80Mtr QRP rig that I built to not only be a usefull
CW QRP rig, but it will produce 1W into a 50 Ohm load "for ever".
I have a Boonton "Microwatter" with a 431-4E "head" that is rated
down to 30MHz. I will check it later today to see if it will work down
to
7MHz, if not I will have to use my Diawa power meter, lowest scale is
5W FS. I have calibrated attenuators so I should be able to reduce
the 1W to 10mw with no trouble. Iwold preffer to move up to so a
frequency higher and further "in" to the ZFSC-2-1 pand pass, but
unless I can fix a ~5W 6mtr CW rig I put away a lifetime ago I will
have to work at 7MHz. Any errors ought to favor the resitive hybrid.

I wonder if my IC-2AT at low power on 146.00MHz will work very well for
a RF test souce, running from a 9V supply, it ought to be able to run
"key down" on low power for a while before overheating.

Terry

In article . com,
wrote:

I will study both pdfs and see if I can come up with a couple of
experiments to see if I understand this.

I am picking up a Pasternak PE2066 DC~6GHz with "N" connectors from
a friend this afternoon.

I have a 1W 7MHz/80Mtr QRP rig that I built to not only be a usefull
CW QRP rig, but it will produce 1W into a 50 Ohm load "for ever". I
have a Boonton "Microwatter" with a 431-4E "head" that is rated down
to 30MHz. I will check it later today to see if it will work down to
7MHz, if not I will have to use my Diawa power meter, lowest scale is
5W FS. I have calibrated attenuators so I should be able to reduce
the 1W to 10mw with no trouble. Iwold preffer to move up to so a
frequency higher and further "in" to the ZFSC-2-1 pand pass, but
unless I can fix a ~5W 6mtr CW rig I put away a lifetime ago I will
have to work at 7MHz. Any errors ought to favor the resitive hybrid.

I wonder if my IC-2AT at low power on 146.00MHz will work very well
for a RF test souce, running from a 9V supply, it ought to be able to
run "key down" on low power for a while before overheating.

Remember to terminate all ports at all times or your results will vary.

The transformer type of splitter will have a "U" shaped response curve
so you need to be within that band for it to work right. The resistive
type is good from DC to some max frequency and you need to be in that
range.

From the transmitter value each output port should be 3 dB down when the
other output port is terminated in 50 ohms. The actual loss will be a
little larger then 3 dB by the splitters insertion loss.

Oh yeah, and your meter is probably a through type instead of a load
type so you need to put a 50 ohm load on the far side of the meter.

When you check each splitter also switch the meter and its load in each
leg to see how balanced the splitter is in each output port.

Be sure to turn the transmitter off before all connections are made or
you could damage your transmitter.

--
Telamon
Ventura, California

Telamon wrote:

Remember to terminate all ports at all times or your results will
vary.

The transformer type of splitter will have a "U" shaped response
curve
so you need to be within that band for it to work right. The
resistive
type is good from DC to some max frequency and you need to be in that

range.

From the transmitter value each output port should be 3 dB down when
the
other output port is terminated in 50 ohms. The actual loss will be a

little larger then 3 dB by the splitters insertion loss.

Oh yeah, and your meter is probably a through type instead of a load
type so you need to put a 50 ohm load on the far side of the meter.

When you check each splitter also switch the meter and its load in
each
leg to see how balanced the splitter is in each output port.

Be sure to turn the transmitter off before all connections are made
or
you could damage your transmitter.

--
Telamon
Ventura, California

I made some measurements last night.
I decided to move from ~7MHz to 146 because 7MHz is just
too close to the lower limit on my mMc splitter.

I have an IC28A that is adjusted to give 1wt in low power.
I used a 10dB Narda 766-10 power attenuator in series with
the splitter input. I used another 766-10 attenuator in series
with the output and the sensor head on Boonton RF power meter.
I used a .5% 50 ohm terminator for the other port

The resitive splitter gave me 6dB of POWER loss.
The MC transformer based splitter gave me just over
3.2dB of loss.

I called a freshly frocked dual PHD, EE and physics, I know
and asked her what the heck is happening. The quick answer
is the resistive splitter is like any other resitive voltage
divider. The transformer unit is like a transformer step down,
or step up. The only losses in a transformer splitter are the
wire losses and the core loss. The R in the MiniCircuits dispates
no power if all ports are matched.

She invited me to "her" lab Friday afternoon and she will use
the modern equipment to verify my measurements.
She suggested that I treat the resistive network and use kichoff's
(?spelling?, it has been too long since college) law to show the
dissapation of each element. It is too late tonight and my head
hurts from this stuff.

Oh, and cctual S-meter measurements show the MC splitter has half
the loss of the resitive splitter.

On a good note, I picked up a 3 section "dB Products" heavy aluminium
mast and another MC splitter like mine and a 1 in 8 out RCA active
LF/HF
splitter rated for use from 100Kc to 40Mc. It used Nuvisters and
appears
to have selectable MB/BCB filters. No manual, and the only diagram is
stincyled on the inside of the lid. And no model number only an old
"Signal Corp" decal. My friend retired ATT microwave engineer and is
moving
to AZ because he is sick of the winter snow here. He has a storage
building full of such junk and I get to cherry pick it and haul it to
another friend.
I hope to check the RCA coupler before going to the university lab.
It would be nice to run a SN and frequency response.

Terry

In article . com,
wrote:

Telamon wrote:

Remember to terminate all ports at all times or your results will
vary.

The transformer type of splitter will have a "U" shaped response
curve
so you need to be within that band for it to work right. The
resistive
type is good from DC to some max frequency and you need to be in that

range.

From the transmitter value each output port should be 3 dB down when
the
other output port is terminated in 50 ohms. The actual loss will be a

little larger then 3 dB by the splitters insertion loss.

Oh yeah, and your meter is probably a through type instead of a load
type so you need to put a 50 ohm load on the far side of the meter.

When you check each splitter also switch the meter and its load in
each
leg to see how balanced the splitter is in each output port.

Be sure to turn the transmitter off before all connections are made
or
you could damage your transmitter.

--
Telamon
Ventura, California

I made some measurements last night.
I decided to move from ~7MHz to 146 because 7MHz is just
too close to the lower limit on my mMc splitter.

I have an IC28A that is adjusted to give 1wt in low power.
I used a 10dB Narda 766-10 power attenuator in series with
the splitter input. I used another 766-10 attenuator in series
with the output and the sensor head on Boonton RF power meter.
I used a .5% 50 ohm terminator for the other port

The resitive splitter gave me 6dB of POWER loss.
The MC transformer based splitter gave me just over
3.2dB of loss.

I called a freshly frocked dual PHD, EE and physics, I know
and asked her what the heck is happening. The quick answer
is the resistive splitter is like any other resitive voltage
divider. The transformer unit is like a transformer step down,
or step up. The only losses in a transformer splitter are the
wire losses and the core loss. The R in the MiniCircuits dispates
no power if all ports are matched.

She invited me to "her" lab Friday afternoon and she will use
the modern equipment to verify my measurements.
She suggested that I treat the resistive network and use kichoff's
(?spelling?, it has been too long since college) law to show the
dissapation of each element. It is too late tonight and my head
hurts from this stuff.

Oh, and cctual S-meter measurements show the MC splitter has half
the loss of the resitive splitter.

On a good note, I picked up a 3 section "dB Products" heavy aluminium
mast and another MC splitter like mine and a 1 in 8 out RCA active
LF/HF
splitter rated for use from 100Kc to 40Mc. It used Nuvisters and
appears
to have selectable MB/BCB filters. No manual, and the only diagram is
stincyled on the inside of the lid. And no model number only an old
"Signal Corp" decal. My friend retired ATT microwave engineer and is
moving
to AZ because he is sick of the winter snow here. He has a storage
building full of such junk and I get to cherry pick it and haul it to
another friend.
I hope to check the RCA coupler before going to the university lab.
It would be nice to run a SN and frequency response.

Well, that's interesting results.

I've used resistive splitters in the past and have gotten the 6dB
division. The resistive splitters I have used in the past are the T and
the delta type. I usually work with scopes so I see 1/2 the voltage or
6dBV on the scope through the splitter and I have no explanation why
you are seeing more than that.

What is the make and model number of the resistive splitter?

--
Telamon
Ventura, California

"Telamon" wrote in message
...
In article . com,
wrote:

Telamon wrote:

Remember to terminate all ports at all times or your results will
vary.

The transformer type of splitter will have a "U" shaped response
curve
so you need to be within that band for it to work right. The
resistive
type is good from DC to some max frequency and you need to be in that

range.

From the transmitter value each output port should be 3 dB down when
the
other output port is terminated in 50 ohms. The actual loss will be a

little larger then 3 dB by the splitters insertion loss.

Oh yeah, and your meter is probably a through type instead of a load
type so you need to put a 50 ohm load on the far side of the meter.

When you check each splitter also switch the meter and its load in
each
leg to see how balanced the splitter is in each output port.

Be sure to turn the transmitter off before all connections are made
or
you could damage your transmitter.

--
Telamon
Ventura, California

I made some measurements last night.
I decided to move from ~7MHz to 146 because 7MHz is just
too close to the lower limit on my mMc splitter.

I have an IC28A that is adjusted to give 1wt in low power.
I used a 10dB Narda 766-10 power attenuator in series with
the splitter input. I used another 766-10 attenuator in series
with the output and the sensor head on Boonton RF power meter.
I used a .5% 50 ohm terminator for the other port

The resitive splitter gave me 6dB of POWER loss.
The MC transformer based splitter gave me just over
3.2dB of loss.

I called a freshly frocked dual PHD, EE and physics, I know
and asked her what the heck is happening. The quick answer
is the resistive splitter is like any other resitive voltage
divider. The transformer unit is like a transformer step down,
or step up. The only losses in a transformer splitter are the
wire losses and the core loss. The R in the MiniCircuits dispates
no power if all ports are matched.

She invited me to "her" lab Friday afternoon and she will use
the modern equipment to verify my measurements.
She suggested that I treat the resistive network and use kichoff's
(?spelling?, it has been too long since college) law to show the
dissapation of each element. It is too late tonight and my head
hurts from this stuff.

Oh, and cctual S-meter measurements show the MC splitter has half
the loss of the resitive splitter.

On a good note, I picked up a 3 section "dB Products" heavy aluminium
mast and another MC splitter like mine and a 1 in 8 out RCA active
LF/HF
splitter rated for use from 100Kc to 40Mc. It used Nuvisters and
appears
to have selectable MB/BCB filters. No manual, and the only diagram is
stincyled on the inside of the lid. And no model number only an old
"Signal Corp" decal. My friend retired ATT microwave engineer and is
moving
to AZ because he is sick of the winter snow here. He has a storage
building full of such junk and I get to cherry pick it and haul it to
another friend.
I hope to check the RCA coupler before going to the university lab.
It would be nice to run a SN and frequency response.

Well, that's interesting results.

I've used resistive splitters in the past and have gotten the 6dB
division. The resistive splitters I have used in the past are the T and
the delta type. I usually work with scopes so I see 1/2 the voltage or
6dBV on the scope through the splitter and I have no explanation why
you are seeing more than that.

What is the make and model number of the resistive splitter?

--
Telamon
Ventura, California

I think you are both in violent agreement - 6dB loss means the output over
the input is 1/2 for voltage (constant impedance), 1/4 for power. A
resistive 2-way minimum loss splitter has 6dB loss; a transformer 2-way
splitter is 3dB.

Tom

In article ,
"Tom Holden" wrote:

"Telamon" wrote in
message

gy.com.. .
In article . com,
wrote:

Telamon wrote:

Remember to terminate all ports at all times or your results
will vary.

The transformer type of splitter will have a "U" shaped response
curve so you need to be within that band for it to work right.
The resistive type is good from DC to some max frequency and you
need to be in that range.

From the transmitter value each output port should be 3 dB down
when the other output port is terminated in 50 ohms. The actual
loss will be a little larger then 3 dB by the splitters
insertion loss.

Oh yeah, and your meter is probably a through type instead of a
load type so you need to put a 50 ohm load on the far side of
the meter.

When you check each splitter also switch the meter and its load
in each leg to see how balanced the splitter is in each output
port.

Be sure to turn the transmitter off before all connections are
made or you could damage your transmitter.

-- Telamon Ventura, California

I made some measurements last night. I decided to move from ~7MHz
to 146 because 7MHz is just too close to the lower limit on my mMc
splitter.

I have an IC28A that is adjusted to give 1wt in low power. I used
a 10dB Narda 766-10 power attenuator in series with the splitter
input. I used another 766-10 attenuator in series with the output
and the sensor head on Boonton RF power meter. I used a .5% 50 ohm
terminator for the other port

The resitive splitter gave me 6dB of POWER loss. The MC
transformer based splitter gave me just over 3.2dB of loss.

I called a freshly frocked dual PHD, EE and physics, I know and
asked her what the heck is happening. The quick answer is the
resistive splitter is like any other resitive voltage divider. The
transformer unit is like a transformer step down, or step up. The
only losses in a transformer splitter are the wire losses and the
core loss. The R in the MiniCircuits dispates no power if all
ports are matched.

She invited me to "her" lab Friday afternoon and she will use the
modern equipment to verify my measurements. She suggested that I
treat the resistive network and use kichoff's (?spelling?, it has
been too long since college) law to show the dissapation of each
element. It is too late tonight and my head hurts from this stuff.

Oh, and cctual S-meter measurements show the MC splitter has half
the loss of the resitive splitter.

On a good note, I picked up a 3 section "dB Products" heavy
aluminium mast and another MC splitter like mine and a 1 in 8 out
RCA active LF/HF splitter rated for use from 100Kc to 40Mc. It
used Nuvisters and appears to have selectable MB/BCB filters. No
manual, and the only diagram is stincyled on the inside of the
lid. And no model number only an old "Signal Corp" decal. My
friend retired ATT microwave engineer and is moving to AZ because
he is sick of the winter snow here. He has a storage building full
of such junk and I get to cherry pick it and haul it to another
friend. I hope to check the RCA coupler before going to the
university lab. It would be nice to run a SN and frequency
response.

Well, that's interesting results.

I've used resistive splitters in the past and have gotten the 6dB
division. The resistive splitters I have used in the past are the T
and the delta type. I usually work with scopes so I see 1/2 the
voltage or 6dBV on the scope through the splitter and I have no
explanation why you are seeing more than that.

What is the make and model number of the resistive splitter?

-- Telamon Ventura, California

I think you are both in violent agreement - 6dB loss means the output
over the input is 1/2 for voltage (constant impedance), 1/4 for
power. A resistive 2-way minimum loss splitter has 6dB loss; a
transformer 2-way splitter is 3dB.

I just calculated it and you are right Tom. I have been a real dufus and
giving Terry a headache for nothing. I generally don't think in terms of
power since I work with small signals.

The resistive splitter is burning half the power as the T types three
16.7 ohm resistors in series are 50 ohms.

Sorry about that Terry. Looks like your transformer type is more
efficient in its bandwidth than the resistive type. Since you measured
3.2 dB on the transformer type you are very close to the ideal splitter
and your measurement on the resistive is also correct at 6 dB so you can
stop scratching you head now.

--
Telamon
Ventura, California

freshly frocked!!??
cuhulin

wrote:
freshly frocked!!??
cuhulin

yawn !!